摘要
云平台正成为IT领域内的一种新的服务热点,大量消耗的能量以及满足用户复杂的需求成为平台应用的巨大挑战,文章提提出了云平台的虚拟机(VM)提供与放置策略,以限制性满足程序对策略求解,提出了一个分层的中间件构架去实现平台的性能与能耗管理策略,各功能模块组合在框架内,整个实现对基础平台与应用透明,容易在线扩展、升级、方便进行适应性调节,实验表明新的方案能在满足用户需求条件下最小化平台能耗,具有使用方便,适应性好等特点,是一种较有理论与应用价值的新方案。
Cloud computing servers is becoming a hot application fields, A key issue is to minimize their cost by maximize their users prof- its. In this article, we propos a strategy of virtual machine provisioning and placement, VM provisioning and placement all are CSP prob- lems, we solve them by mature programming module, To realize the strategy a middleware architecture is proposed, all strategy module is organized here, it is transparent for users. Experiments show the new strategy can low energy consumption and meet users SLA constraint, we also modify modules in middleware architecture to suit for new users demands, and tunable the parameters to suit for new application surrounds, it is a future method.
出处
《电脑知识与技术》
2012年第3期1639-1643,共5页
Computer Knowledge and Technology
基金
高校博士点学科专项基金(20080560019)
湖南省教育厅项目(11C0698基于低能耗的海量存储系统理论与关键技术研究)
湖南人文科技学院校级青年基金“智能存储系统I/O研究”(2008QN002)
关键词
虚拟机管理
云平台
能耗
用户需求
virtual machines management
cloud infrastructure
power consumption
user demands